On Friday, August 3, 2018 at 9:33:03 PM UTC+3, vdelecroix wrote:
>
>
>
> On 31/07/2018 05:57, Erik Bray wrote:
> > On Mon, Jul 30, 2018 at 8:34 PM Jeroen Demeyer <j.de...@ugent.be
> <javascript:>> wrote:
> >>
> >> On 2018-07-30 18:31, Erik Bray wrote:
> >>> This is a bit far-out, but what if we just didn't show tracebacks at
> >>> all by default?
> >>
> >> Sorry to say, but that would be a horrible idea. Even if the tracebacks
> >> are completely useless for ordinary users, they allow developers to
> >> debug an issue when posted on a mailing list/asksage. I would avoid
> >> anything that makes it harder for users to get help from developers.
> >
> > I agree but if you read the rest of my message you would see that I
> > also suggested making it very easy to obtain the actual traceback
> > (even including instruction for doing so with every error). That's no
> > harder--easier even--than what we already ask of users when give them
> > a path to the error log when Sage crashes on startup.
> >
> > To be concrete, instead of seeing:
> >
> > sage: 1 / 0
> >
> ---------------------------------------------------------------------------
> > ZeroDivisionError Traceback (most recent call
> last)
> > <ipython-input-10-b43c99900d8f> in <module>()
> > ----> 1 Integer(1) / Integer(0)
> >
> >
> /home/embray/src/sagemath/sage-python3/local/lib/python3.6/site-packages/sage/structure/element.pyx
>
>
> > in sage.structure.element.Element.__truediv__
> > (build/cythonized/sage/structure/element.c:13020)()
> > 1729 cdef int cl = classify_elements(left, right)
> > 1730 if HAVE_SAME_PARENT(cl):
> > -> 1731 return (<Element>left)._div_(right)
> > 1732 if BOTH_ARE_ELEMENT(cl):
> > 1733 return coercion_model.bin_op(left, right, truediv)
> >
> >
> /home/embray/src/sagemath/sage-python3/local/lib/python3.6/site-packages/sage/rings/integer.pyx
>
>
> > in sage.rings.integer.Integer._div_
> > (build/cythonized/sage/rings/integer.c:13584)()
> > 1952 """
> > 1953 if mpz_sgn((<Integer>right).value) == 0:
> > -> 1954 raise ZeroDivisionError("rational division by zero")
> > 1955 x = <Rational> Rational.__new__(Rational)
> > 1956 mpq_div_zz(x.value, self.value, (<Integer>right).value)
> >
> > ZeroDivisionError: rational division by zero
> >
> >
> > users would see:
> >
> > sage: 1 / 0
> > ZeroDivisionError: rational division by zero (enter %tb to see the
> > full error traceback)
>
> +1 here, too.
I'd even have
> ZeroDivisionError: rational division by zero (enter %tb to see the
> full error traceback - important for reporting errors!)
I am +1 for this having this kind of behavior. There should also be
> command line arguments to enable/disable this (perhaps --with-traceback
> and --without-traceback). As for the default I don't really care since
> the "sage" command on the system can be made anything.
>
> > As William noted this is already roughly the behavior on SageNB, so I
> > would wager that it's not a "horrible idea".
>
> I would even call it a "good idea"!
>
> Vincent
>
--
You received this message because you are subscribed to the Google Groups
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.
